ArticleNature2026
A benchmark of expert-level academic questions to assess AI capabilities.
Center for AI Safety et al.PubMed ↗Full text ↗Publisher ↗
No numbers read from the abstract.
9 papers cite it
this papercites it
ArticleNature2026
Center for AI Safety et al.PubMed ↗Full text ↗Publisher ↗
No numbers read from the abstract.